Fix ZIP Filenames
ZIPs made on Windows in another language often show file names as gibberish such as ╩╓╗·.txt on other computers. This tool detects the original encoding (Shift_JIS, GBK, Big5, Korean, Cyrillic and more), shows you the corrected names, and rewrites the ZIP with proper UTF-8 names. File contents are copied unchanged.
Fix ZIP Filenames
Detect the language a ZIP's file names were written in and rewrite them as UTF-8. File contents stay untouched.
Drop the ZIP with garbled file names
or choose one from your device
Files stay on your device. Nothing is uploaded.
How to use Fix ZIP Filenames
- 1
Open the ZIP
Drop the archive whose file names look like gibberish, or choose it with the file picker.
- 2
Check the detected encoding
Names stored without the UTF-8 flag are scored against every supported encoding, and the most likely one is preselected.
- 3
Compare before and after
Sample names are shown as they look now and as they will look after the fix. If they still look wrong, pick another encoding.
- 4
Download the fixed ZIP
A new ZIP is written with the corrected names stored as UTF-8. File data is copied byte for byte.
Why use Aback ZIP's Fix ZIP Filenames?
Japanese, Chinese, Korean, Cyrillic and more
Shift_JIS, EUC-JP, GBK, GB18030, Big5, Korean CP949, CP866, Windows-1251, KOI8-R, CP437 and the Windows code pages for Central European, Western, Greek, Turkish, Hebrew, Arabic, Thai and Vietnamese.
Preview before you commit
The most likely encoding is chosen from the raw bytes, and before/after samples let you confirm it or try another one.
No password needed
Only the names change. File data, including encrypted files, is copied byte for byte, so a password-protected ZIP can be fixed without the password and stays protected.
Names that read the same everywhere
Fixed names are stored as UTF-8 with the UTF-8 flag set, so other systems no longer have to guess. The archive is processed in your browser and not uploaded.
Common use cases
Japanese file names on a Mac
A ZIP from a colleague in Japan shows unreadable names on your Mac. Re-read them as Shift_JIS and something like âeâXâg.txt becomes テスト.txt again.
Chinese archives from Windows
Names such as ╬─╝■.txt usually mean GBK from a Simplified Chinese Windows PC, or Big5 for Traditional Chinese.
Russian names on Linux
Archives made with Windows Explorer on a Russian system often store names in CP866, which unzip and other tools show as accented letters and symbols.
International client deliveries
Fix the names once and pass on an archive that opens correctly for everyone, instead of asking the client to re-send it.
Old archives from CDs and backups
ZIPs made before UTF-8 names were common often use a legacy code page. Restore readable names before you extract or archive them again.
Before converting or merging
Fix the names first so Merge ZIP, ZIP to TAR or Extract ZIP carry readable names into the result.
About Fix ZIP Filenames
Why names turn into gibberish
The ZIP format did not originally record which character set a file name uses. Many Windows tools, including Windows Explorer's built-in Send to > Compressed (zipped) folder and older WinRAR and 7-Zip settings, store names in the computer's legacy code page: Shift_JIS on a Japanese system, GBK on a Chinese one, CP866 on a Russian one. They don't mark it anywhere.
A computer with the same language setting guesses correctly, so the sender sees nothing wrong. Everywhere else, the bytes are read with a different code page and the names come out as accented letters, Greek letters and box-drawing symbols. ZIP has a flag (general purpose bit 11) that marks names as UTF-8, and this tool sets it on every name it fixes.
Names that were already damaged when the ZIP was created, for example with characters replaced by ? or _, can't be repaired. The original characters are not in the file, so no encoding brings them back.
Which encoding to choose
| Language of the names | Encodings to try |
|---|---|
| Japanese | Shift_JIS (CP932), then EUC-JP |
| Chinese (Simplified) | GBK (CP936), then GB18030 |
| Chinese (Traditional) | Big5 (CP950) |
| Korean | CP949 / EUC-KR |
| Russian and other Cyrillic | CP866, Windows-1251, KOI8-R |
| Central European | Windows-1250 |
| Western European | Windows-1252, CP437 |
| Greek, Turkish, Hebrew, Arabic, Thai, Vietnamese | Windows-1253, 1254, 1255, 1256, 874 and 1258 |
| Any language, from a tool that wrote UTF-8 without the flag | UTF-8 (without flag) |
You rarely need the table: the tool scores every encoding against the raw name bytes and preselects the best match. Use it when the preview still looks wrong, for example Big5 instead of GBK for a Traditional Chinese archive.
What changes and what doesn't
- Only names stored without the UTF-8 flag that contain non-ASCII bytes are re-read. Names already marked as UTF-8 are left alone.
- File data, including encrypted data, is copied byte for byte. Nothing is decompressed or recompressed.
- The archive comment is kept.
- If no name in the archive is legacy-encoded, the tool tells you there is nothing to fix instead of writing an identical copy.
What is kept and what changes
| Item | Result | Notes |
|---|---|---|
| File names | Changed | Re-read with the chosen encoding and stored as UTF-8 with the UTF-8 flag. |
| Names already marked UTF-8 | Not changed | |
| File contents | Preserved | Copied byte for byte, without recompression. |
| Encrypted files | Preserved | Still encrypted with the same password; no password is needed. |
| Archive comment | Preserved | |
| Archive file | Rebuilt | A new ZIP; your original is not modified. |
Limitations
- Names that were damaged when the ZIP was made, for example with characters replaced by
?, can't be restored. - One encoding is applied to every name that needs fixing. An archive that mixes names from two different legacy code pages can't be fully fixed in one pass.
- Detection is a best guess from the bytes. Short names, and Western code pages that look alike, may need you to pick the encoding from the preview.
- Only file names are converted. Text files inside the archive keep their original encoding.
- Only ZIP archives are supported.
Processing happens locally in your browser. Your files are not uploaded, and closing the tab discards everything.
Related tools
Fix ZIP Filenames: frequently asked questions
No. The archive is read and rewritten in your browser, and the fixed ZIP is created on your device. Nothing is sent to a server and no account is needed.
The ZIP was most likely made on Windows with a non-English language setting, which stores names in a legacy code page such as Shift_JIS or GBK without saying so. Your computer read them with a different code page. The guide Why ZIP file names are garbled explains it in more detail.
Usually you don't have to: the most likely encoding is preselected. Check the before/after samples. If they still look wrong, try the other encodings for the language you expect, such as EUC-JP for Japanese or Big5 for Traditional Chinese.
No. Only the names change. File data is copied byte for byte, so every file extracts exactly as it did before.
Yes, without the password. The encrypted data is copied as it is, so the files stay protected with the same password.
All names are already marked as UTF-8 or contain only plain ASCII. If they still look wrong, they were probably damaged by the program that made the ZIP, and the original characters are no longer in the file.
Use an archiver that stores names as UTF-8 with the UTF-8 flag. Create ZIP always does this.